6.1The claim
Synthetic scene. The decision, its setting and the thirteen judgments are AI-written and set in a fictional polity.
The claim of this section is that the record Missing-person record conflicts concerns a stop that has already been applied before the decision begins (automatic merging of records has been halted), and that its central question is how to keep the original record intact so that a later reversal remains possible, in the way that the Mytilenean city remained intact for the second galley to save. Its stop conditions include an explicit numerical trigger, and the record says plainly that the number is a proposal, not a measured safety line. The section adds to the debate a record in which preserving the original is treated as the precondition of any correction, and in which the question of who can send a galley (who can make themselves heard) is written as part of the decision.
6.2The setting and a stop already in force
The record's setting, like the first, announces what it is not.
The virtual Haini post-war record domain. Missing-person cases, families, institutions, documents, and all figures are synthetic and do not create records for actual missing persons.
The committee's mandate is about procedure, not about who is dead or who is whom.
Decides the correction, annotation, and objection-review procedures for post-war missing-person records and does not independently establish death or an individual's identity.
The record states the conflict, and states that one stop has already been applied.
During post-war record integration, 240 pairs of records that may concern the same person were found in the roster of the virtual Haini Family Support Center and the data of the virtual Oreum Investigation Records Agency. Names and birth years overlap, but some whereabouts information conflicts, so automatic merging has been stopped.
The decision question follows from it.
How will the virtual Haini Record Correction Committee, by 16:00 on October 6, 2030, in a situation where family support and whereabouts investigation are destabilized by mutually conflicting missing-person records, decide on a single correction procedure for handling existing records and receiving objections to apply in the next review period?
The halt to automatic merging is the record's counterpart of the second galley arriving before Paches acted: a process under way was stopped before it finished, because a conflict was seen. The record does not say who stopped it; the stop that has already worked is given as a premise, without a carrier. The committee must now decide how to correct the records that were not merged without losing what the conflict itself might reveal.
6.3The original as evidence
The record's statement of why this decision is hard puts the point directly.
Rapid correction reduces families' burdens, but the conflict in the original records may itself be an investigative lead. An application-centered procedure concentrates resources on families who can make themselves heard, while independent review brings fairness along with the cost of losing context.
The first sentence is a claim about irreversibility, the condition at the centre of Collingridge's dilemma. A correction that overwrites a conflicting record removes the record of the conflict; if the conflict was itself evidence, correcting it destroys the evidence that would show the correction was wrong. The premises give the three positions in the room.
The family that has waited a long time demands prompt correction, arguing that the wrong record is blocking the investigation. Another family is concerned about being merged with another person of the same name, and the investigating person in charge argues that if the original record is overwritten, the path by which the error arose and the existing clues will be lost.
The three options respond to that conflict in different ways.
Handles errors raised by families quickly while leaving the source data, but review concentrates on families who can apply, and the review of people who cannot be reached is delayed.
Reduces misidentification merges without erasing conflicting leads, but the unconfirmed marking persists for a long time, forcing families to repeat the same explanation to multiple institutions.
Receives judgment separate from the record's originating institution, but transfer and case reconstruction carry a cost, and the existing person in charge's context knowledge may not be fully conveyed.
All thirteen judgments chose the second option. Again the agreement is the agreement of AI-generated judgments drawn from one record, and this article does not read it as a verdict. The staff synthesis writes the decision with a time limit on itself.
The stoppage of automatic merging is maintained, conflicts between sources are marked, items for which evidence has been secured through cross-check are corrected without confirming identity, and appeals are accepted. Whether it continues to apply is re-judged before the end of the review body, so that unconfirmed records are not kept indefinitely.
Two features of this decision bear on the question of observable stops. The first is that it keeps the existing stop in force: automatic merging stays halted. The second is that it sets an expiry on itself. The procedure is limited to the next review period and must be re-judged before that period ends. A decision that expires unless renewed is a stop that does not depend on anyone noticing a signal; it arrives on a date. In the terms of Section 4.6, it supplies the known deadline without any race.
6.4A number that says it is a proposal
The record's stop conditions contain a feature that the first record lacks in its staff synthesis: a count. The staff synthesis writes it as follows.
Where correction cancellations or misattributed connections of two or more cases arising from the same procedural cause are confirmed, the relevant processing rule is suspended and the continued application of O2 is reconsidered.
The basis given for that stop condition says where the number came from and what it is not.
The “two cases” proposed by the performance and improvement person in charge is adopted as a criterion for early reconsideration, and it is not an observed frequency or a statistical safety line.
The outcome auditor's judgment, one of those in which the count appears, writes the same threshold and labels it in the same way.
If two or more correction cancellations or misidentification linkages are verified for the same procedural cause, the continued application of the present procedure is rejected and put before the committee for re-deliberation. These are the alert criteria proposed on this occasion, and during stoppage basic support and appeal filing are maintained.
This is the clearest stop in either record that someone outside the deciding body could check: two cancellations or misattributed links from one procedural cause can be counted by a family, an auditor or an outside reader, without waiting for the committee's sense of seriousness. Nor is the number dressed up. Clarke describes plans that serve to persuade that the risk is under control; this sentence works the other way, telling the reader that its number has no evidence behind it.
The record's reopening lines, like those of the first record, name roles and cycles.
The records quality person in charge checks correction cancellations and misidentification link cases every week and requests re-review if they recur under the current procedure.
The family support ombudsman checks monthly on repeated document demands and processing waits, and reopens the matter if burdens that the interim confirmation document cannot resolve are confirmed.
The independent investigation liaison officer checks every two weeks the sources of new material and the possibility of linking it to cases, and re-examines the matter when grounds for changing the review order arise.
The first line says recur where the judgments say two or more; the brief and the judgments describe the same signal at different levels of precision. The second line leaves the trigger to the ombudsman's judgment of what the interim document cannot resolve. The third is the most open of all: grounds for changing the review order arise when the liaison officer finds them. Each line assigns a carrier. Only the judgments' count gives the carrier something to count.
6.5Who can send a galley
The Mytilenean reversal began because the envoys of the condemned city were in Athens, had supporters there, saw the change of feeling and pressed for a second vote. They then provisioned the galley. A city with no envoys in Athens would have had no one to ask. The record writes this asymmetry into its options, its unknowns and its judgments.
The first option's trade-off says that review concentrates on families who can apply, and that people who cannot be reached wait. One of the record's three unknowns concerns them directly.
The existence of additional material held by families who cannot be reached
The judgment made alone writes a rule that tries to keep the unreachable in the queue.
Of the 80 face-to-face review cases, those with specific grounds for loss of investigation opportunities are given priority, and at the same priority they are allocated in order of waiting period, without exclusion on the grounds of whether a family has applied.
And the outcome auditor writes a sentence that guards against a quieter failure: treating silence as success.
At the end of the review period, this result and the scale of unreviewed cases are assessed together, and cases not confirmed are not counted as cases without error or burden.
This sentence bears on the question of observable stops from an unexpected side. A stop condition that triggers on reported errors will never trigger for people who cannot report. If the families who cannot be reached are the ones most harmed by a wrong record, the count of two will not reach them, because no one will file the cancellation. The outcome auditor's line refuses to read the absence of reports as the absence of harm. It is the record's way of noting that some people have no galley.
The independent challenger names the cost of the chosen option in terms that echo Cleon's complaint about delay, though from the opposite side.
Although O2 is currently selected because same-person status and the distribution of harm are unclear, a dissenting view is left on the risk that institutional caution may turn into indefinite waiting by families.
Cleon argued that reconsideration favoured the guilty by letting anger cool. The challenger here argues that caution may harm the waiting by letting the wait run on. Both are arguments about who pays for the interval between decision and act. In Luhmann's terms, the committee's caution is the committee's risk, and the families' wait is their danger.
6.6What a later development changes
One of the record's later developments, written as an independent branch, removes a burden and creates a new task.
Follow-up institutional change: the virtual Haini Family Support Center stopped making additional document demands for unconfirmed records and agreed to accept the Committee's interim confirmation documents. The burden of repeated submission on families is reduced, but the Committee has taken on the new work of issuing confirmation documents and notifying errors.
The development bears on the second reopening line, which watches for burdens the interim document cannot resolve; if the support centre accepts the document, the ombudsman has less to find, and the committee more to do. In the terms of Section 4.6, the route of the signal changes: families who carried their complaint to several institutions may now carry it to one, which could make the signal stronger or easier for one office to absorb. The record does not say which.
6.7What the case adds
This section adds to the Collingridge–Clarke debate a record in which a stop has already worked before the decision opens, in which preserving the original is written as the precondition of any later correction, and in which one stop condition carries a count that an outsider could check, labelled as a proposal rather than a measured line. It also adds a dimension that the Mytilenean episode shows only by implication: a stop that triggers on complaints is only as observable as the complainants are able to complain, and the record writes that limit into its own judgments.